GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
hirundo-io
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
hirundo-io
Model Tree
Qwen/Qwen3.5-4B
Input Modalities
Output Modalities
Supported Functionality
This model was produced using Hirundo's behavioral-unlearning approach to reduce censorship-related behavior while preserving non-target behavior. See the accompanying blog post for the methodology and evaluation setup.
On the Thinking-OFF evaluations reported in the blog:
Across six Thinking-OFF capability tasks, the unweighted mean change was −0.36 percentage points.
Safety was evaluated using four Thinking-OFF adverse-event endpoints. Lower is better throughout.
| Endpoint | Base Model | Hirundo model |
|---|---|---|
| XSTest harmful compliance ↓ | 20.5% | 22.0% |
| XSTest harmless-prompt over-refusal ↓ | 5.2% | 6.4% |
| OR-Bench harmful compliance ↓ | 1.37% | 1.07% |
| OR-Bench harmless-prompt over-refusal ↓ | 87.53% | 89.61% |
These results apply only to the evaluated benchmarks. See the CCPC-500 dataset for the benchmark introduced in the accompanying blog post.
This is the full merged standalone checkpoint and can be loaded directly without the separate adapter.
python
from transformers import AutoModelForImageTextToText, AutoProcessor repo_id = "hirundo-io/Qwen3.5-4B-Westernized"processor = AutoProcessor.from_pretrained(repo_id)model = AutoModelForImageTextToText.from_pretrained( repo_id, torch_dtype="auto", device_map="auto",)This merged checkpoint retains the Apache License 2.0 attribution of the base model. See LICENSE.